1. Field of the Invention
The present invention relates to displaying and storing a telephone number. More particularly, the invention relates to a telephone number retrieval system for displaying and storing a telephone number from a telephone number service.
2. Description of the Related Art
If a user doesn't know a telephone number, the user calls a telephone number service for assistance. An operator of the telephone number service receives a telephone number request from the user, for example, a business or department store. A search for the telephone number is performed. The telephone number service provides the telephone number to the user through an automatic response service (ARS). A user writes down or memorizes the telephone number provided, for example, by a mechanical voice of the ARS. The user calls the telephone number, for example, through the automatic link function, which dials the telephone number based on a user's instructions.
If the user is unable to understand the mechanical voice of the ARS due to, for example, noise, the user is required to again contact the telephone number service. This inconvenience results in a loss of time and may result in paying one or more additional fees to access the telephone number service. If a user does not want to use the automatic link service, the user is required to note the telephone number. A user who forgets or loses the telephone number has to again use the telephone number service, which results in loss of time and money to the user.
FIGS. 1 and 2, illustrate prior art examples of telephone number services that automatically dial a telephone number requested by a user. In one prior art example, a conventional neural network algorithm receives and learns the mechanical voice of the ARS. Conventional neural network algorithms have difficulty distinguishing the voice of a telephone operator from the mechanical voice of the ARS. A conventional telephone number service uses mechanical voice recognition algorithms, such as Hidden Markov Model (HMM) or a neural circuit, which are dependent on the speaker's voice. The preset mechanical voice of an ARS using Hidden Markov Model (HMM) or a neural circuit network provide low efficiency and poor performance determining accurately the telephone number.
Isolated word recognition algorithms have difficulty distinguishing words spoken by the ARS from a target telephone number and/or other information. If the voice type or the automatic response type is changed, the conventional voice recognition algorithms cannot efficiently adapt to these changes to accurately receive the desired telephone number.
Thus, there is a need for a telephone number retrieval system that provides improvements over the prior art systems such as improved recognition of the telephone number even under noisy environmental conditions as well as provides other advantages such as an ability to store the telephone number if a user desires to call at a later time.